What is the optimal management of pediatric nontuberculous mycobacterial cervicofacial lymphadenitis?

Nontuberculous mycobacterial lymphadenitis (NTML) is a common cause of subacute and chronic cervicofacial masses in children under 5 years of age (Fig. 1).1 The nontuberculous mycobacteria (NTM) are environmentally ubiquitous organisms that tend to infect young, immunocompetent children following ingestion of contaminated material.1 Most clinical infections are caused by Mycobacterium avium.2, 3 Tuberculin skin test and acid-fast staining may suggest mycobacterium as the etiologic agent, which may be further distinguished by blood-based testing (interferon-gamma release assays) or polymerase chain reaction (PCR) and culture, although reported yield on culture data is roughly 50% (Fig. 2).1, 2 The natural history of disease involves slowly enlarging lymphadenopathy of the cervicofacial region, most commonly in the anterior submandibular triangle, cervical jugular chain, or periparotid region.1 Most children with NTML have no systemic symptoms, and infection generally does not extend beyond the local site.2 As the disease progresses, there may be increasing fluctuance or overlying violaceous skin discoloration to the involved lymph node region.1, 2 In late disease, children may develop a chronically draining fistula.2 Current management options include observation, directed antibiotic therapy, fine needle aspiration (FNA), incision and drainage (I&D), curettage, and complete surgical excision. Primary or adjuvant clarithromycin and rifampin have demonstrated some efficacy against NTM, requiring a minimum duration of 6 to 8 weeks but often extending much longer with varying success rates.4 Most reports, including infectious disease literature, favor complete surgical excision of NTML, with antibiotic therapy reserved for children in whom surgical excision is incomplete or who recur.2, 4 Numerous studies of pediatric NTML describe management options including observation, antibiotic therapy, and incomplete or complete surgical excision of the involved NTM lymph nodes. The morbidity of any treatment must be weighed against that of the disease's untreated, natural course. In a 2008 retrospective study, Zeharia et al. followed 92 children with culture-positive NTML whose parents chose the option of “observation only.”3 At initial presentation, all subjects had unilateral lymphadenitis; 25% had skin discoloration; and none had a draining fistula.3 Complete disease resolution was achieved in all patients; however, this required up to 12 months in some cases.3 During the observation period, 97% of children developed a fistula overlying the affected lymph node that drained purulent material for 3 to 8 weeks before clearing.3 Scarring was reported at fistula sites; otherwise, no adverse events were noted at the 2-year follow-up.3 Antibiotic regimens for NTML have repeatedly demonstrated inferior cure rates compared with surgical excision.1, 2, 4 Whereas tuberculous mycobacteria respond well to anti-tuberculous drugs, NTM demonstrate little clinical response.2 Furthermore, with the exception of macrolides, in vitro susceptibility testing has not been found to correlate with in vivo therapeutic response, leading some authors to consider routine susceptibility testing to be unnecessary in most clinical situations.2, 5 The benefit of antibiotic therapy over watchful waiting for management of NTML in immunocompetent children is therefore unclear. For example, a 2011 randomized controlled trial (RCT) comparing clarithromycin and rifabutin antibiotic therapy with a wait-and-see policy demonstrated no statistical difference in time to disease resolution (36 weeks with antibiotics vs. 40 weeks without).5 Antibiotics may, however, reduce the rate of fistula formation when compared to observation only. Lindeboom et al. showed that 44% of patients’ bulky lymphadenopathy regressed without fistula formation with 12 weeks of clarithromycin and rifabutin, compared to Zeharia et al.'s 97% fistula rate with observation only.3, 4 Adverse effects and costs, however, may make antibiotics less preferred.5 Lindeboom et al. found that 74% of the antibiotic arm reported adverse effects compared to 20% in the surgical arm.4 Adverse events, most commonly reversible tooth discoloration, fever, and fatigue, were reported in 67.8% of patients treated with anti-mycobacterial antibiotics in the meta-analysis.2 In immunocompetent children, the literature shows that antibiotics not only yield inferior cure rates and prolong treatment course when compared to surgical excision but may also confer additional adverse effects.1, 2, 4, 5 The treatment of NTM with the highest proven eradication rate (92%–98%) is complete surgical excision of the node(s) and/or involved skin.1, 2, 4 In 2007, Lindeboom et al. randomized 100 pediatric patients with culture- or PCR-proven NTML to either surgical excision or antibiotic therapy.4 Surgical patients received one dose of preoperative flucloxacillin and underwent modified neck dissection. Patients in the antibiotic therapy arm received clarithromycin and rifabutin for 12 weeks.4 The authors reported a 96% cure rate for the 50 surgical patients versus a 66% cure rate for the 50 patients who received antibiotics.4 In 2015, Zimmerman et al. published a systematic review and meta-analysis of 1,951 children with NTML.2 They reported an adjusted mean cure rate of 98% (95% confidence interval [CI] 97.0%–99.5%) for complete excision, 73.1% (95% CI 49.6%–88.3%) for antibiotic therapy, and 70.4% (95% CI 49.6%–88.3%) for observation.2 Complete surgical excision, however, is not without risks. In addition to potential morbidity of any surgical procedure requiring general anesthesia, young children with high skin elasticity are prone to scar widening following the excision of infected skin and tissue.1-4 Furthermore, nodal inflammation and overlying dermal necrosis may complicate dissection of facial nerve branches, particularly near the inferior edge of the mandible or intraparotid.1, 4 Indeed, marginal mandibular nerve injury accounted for all seven cases (14%) of postoperative facial nerve weakness in Lindeboom et al.'s 2007 RCT; all but one child regained full function within 12 weeks (2%).4 Even with intraoperative facial nerve monitoring, temporary facial nerve weakness has been reported in up to 20% of cases, with permanent nerve injury occurring in roughly 2% of children subjected to nerve dissection.2, 3 Less-invasive surgical interventions such as FNA, I&D, and curettage pose less risk to the facial nerve but are less effective and often require additional procedures to achieve cure.2 Indeed, Zimmerman et al. showed a pooled cure rate of 37.5% for FNA, 34.2% for I&D, 61.2% for curettage, and 88.5% for complete surgical excision.2 This meta-analysis demonstrated that, compared to no intervention, complete surgical excision was the only intervention that significantly led to cure (odds ratio 33.3, P value <0.0001).2 All other interventions, including incomplete excision, curettage, I&D, FNA, antibiotics, or other surgical combinations, did not demonstrate statistically significant improved outcomes and in some cases demonstrated statistically worse outcomes.2 Spinelli et al. report a cure rate of 97.7% for patients who underwent complete surgical excision versus 88.2% for patients who underwent I&D with and without curettage (P = 0.06).1 In fact, nonexcisional surgical treatment often leaves patients prone to fistula formation.1, 2 Complete surgical excision should be offered primarily because this offers the most expedient and definitive cure for NTML. In cases for which complete surgical excision is deferred due to concern of permanent facial nerve dysfunction or scarring, a prolonged course of antibiotics may reduce the likelihood of draining fistula formation but is unlikely to accelerate time to resolution. The data reviewed derive from level 2 and 3 evidence, including two RCTs, one meta-analysis, and two large retrospective analyses. The authors would like to thank H. Cody Meissner, MD for his input in the preparation of this manuscript.